Mushrooms are food commodities that are nutritious for the body. Even mushrooms can be a cure for certain diseases. But not all types of mushrooms are nutritious for the body, there are types of mushrooms that can even be bad for the body or toxic. Therefore the classification of edible and poisonous mushrooms is very important in order to be able to consume the right mushrooms. The classification method used to classify edible or poisonous mushrooms is Naive Bayes with feature selection based on Association Rule. Prior to classification using Naive Bayes, an Association Rule-based feature selection is performed by selecting features in the rule that meet minimum support and minimum confidence. The best accuracy result of mushrooms classification with Naive Bayes and feature selection is 95% with one selected feature. Meanwhile, the seven selected features produce an accuracy of 94%. If without feature selection the resulting accuracy is 95%. Although the accuracy with feature selection is not better than without feature selection, by using feature selection the computational performance of the model becomes more efficient and accuracy only decreases by 1%. This means that feature selection based on Association Rule and classification using Naive Bayes is successful in classifying mushrooms.
Copyrights © 2022